Diffusive clustering of interacting Brownian motions on \(\mathbb Z^ 2\). (Q1877535)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Diffusive clustering of interacting Brownian motions on \(\mathbb Z^ 2\).
scientific article

    Statements

    Diffusive clustering of interacting Brownian motions on \(\mathbb Z^ 2\). (English)
    0 references
    0 references
    7 September 2004
    0 references
    Long-time behaviour of interacting diffusions which are the unique strong solution to a system \[ dx_ {t}(i) = \sum _ {j\in \mathbb Z^ 2} a(i,j)\bigl (x_ {t}(j) - x_ {t}(i)\bigr )\,dt + dW_ {t}(i), \quad i\in \mathbb Z^ 2, \] of coupled stochastic differential equations is studied. It is assumed that \(W(i)\), \(i\in \mathbb Z^ 2\), are independent standard one-dimensional Wiener processes and \(a\) is a kernel of an irreducible non-degenerate random walk on \(\mathbb Z^ 2\) with finite \((2+\delta)\)th moments for some \(\delta >0\). Let us consider rescaled block averages defined by \[ \theta _ {t}(\alpha ) = \left ({2\pi \sqrt {\det Q}\over \log t}\right ) ^ {\!\! 1/2} \frac 1{| B(t^ {\alpha /2})| }\;\sum _ {i\in B(t^ {\alpha /2})} x_ {t}(i), \quad t>0, \;\alpha \in [0,1], \] where we set \(B(r) = \{i\in \mathbb Z^ 2;\; \| i\| _ \infty \leq r\}\) and \(Q\) is a \((2\times 2)\)-matrix defined by \(\langle Qy,y\rangle = \frac 12\sum _ {k\in \mathbb Z^ {2}}(a(k,0) + a(0,k))\langle k,y \rangle ^ 2\), \(y\in \mathbb R^ 2\). Suppose that the initial condition \(x_ 0\) is ergodic and has a finite moment of order \(2+\delta \). Let \(W\) be a Wiener process. Then it is shown that the process \((\theta _ {t}(\alpha )\), \(\alpha \in [0,1])\) converges in law as \(t\to \infty \) in the Skorokhod topology to \((W_ {1-\alpha }\), \(\alpha \in [0,1])\).
    0 references
    0 references
    0 references
    interacting diffusions
    0 references
    pathwise convergence
    0 references
    long-time behaviour
    0 references
    Gaussian processes
    0 references